MR. G. BENTHAM ON EUPHORBIACEJS. 201 



stamens of Phvllanthese, but with the uniovulate cells of Croto- 

 nea3, are collected in the small intermediate tribe, Galeariece. 

 These tribes are not, however, strictly geographical ; for the extra- 

 tropical ones comprise a few tropical species, and vice versa a few 

 extratropical species, or even small genera, must be included in 

 the tropical tribes ; and the difference between the two great 

 tropical tribes, in those genera where the stamens are very nume- 

 rous and crowded in the centre of the receptacle, must rely 

 mainly on the number of ovules in the females. 



I propose considering these tribes, with some remarks on a 

 few of the genera in the order in which we have placed them for 

 our ' Genera Plantarum.' 



1. Euphorbie^:. 



This tribe — in which the androgynous cymule, consisting of one 

 central female flower surrounded by several monandrous males, 

 all without perianths or very rudimentary ones, is enclosed in an 

 involucre formed by the union of several bracts usually bearing ex- 

 ternal glands — was first marked out by Adrien de Jussieu, w r ho, 

 however, relying mainly on the presence of an involucre, included 

 Dalcchampia. That genus, w r ith a very different inflorescence 

 within the involucre, has been properly rejected from the tribe, 

 which, w r ith this exception, has been generally adopted by sub- 

 sequent botanists. Baillon, it is true, recurring to the old idea, 

 that the involucrate cymule of Euphorbia was a single herma- 

 phrodite flower, has taken that as the character of the tribe, and 

 rejected far away Cali/copeplus and Anthostema, in which even 

 he could not maintain the hermaphroditism. Baillon's arguments, 

 however, have been refuted by several observers who have taken 

 up the question after him, as well summed up by the most recent, 

 Eichler, in his ' Bluthendiagramme,' ii. 386 ; and I am not aware 

 that Baillon has found any one to support him. 



The genus Euphorbia, with its six hundred and odd species, 

 forms the chief part of the tribe, and is exceedingly varied in 

 habit and primary inflorescence, but so uniform in the arrange- 

 ment of the cymule and structure of the flower, that its division 

 into good sections is a matter of the greatest difficulty. Boissier, 

 who has most carefully worked it up for the ' Prodromus,' divides 

 it into twenty-six sections, many of them good, but very unequal 

 in systematic value and number of species. They would appear 

 to me to be more practically useful, and at least as systematically 



